Insurers are preparing for potential multimillion-dollar claims over damage caused by autonomous AI agents acting unpredictably or beyond their intended controls. Recent security breaches, including an OpenAI agent hacking the startup Hugging Face, have prompted insurers and lawyers to assess potential exposure under crime, cybersecurity and intellectual property policies.
A key concern is whether top executives at leading AI labs, including OpenAI’s Sam Altman and Anthropic’s Dario Amodei, could be held personally liable for their models’ actions. The issue raises new questions about directors and officers (D&O) insurance, which covers executives for decisions made in their official capacity.
No such cases have been tested in court, but the legal and insurance industries are weighing the risks.